Learn / Core arithmetic
What the edge number actually means.
PICKD does not predict player performance. It compares two published quantities at the exact same prop line: the probability implied by a payout and a fair probability derived from a two-sided reference market.
1. Find the payout’s break-even.
For an all-or-nothing slip with total payout multiplier M and n independent legs, the per-leg break-even is:
This is the hit rate each leg would need if every leg had the same probability. Book-specific multipliers can imply different break-even probabilities for each side.
2. Remove the reference margin.
A two-sided book publishes an over price and an under price. Their raw implied probabilities usually add to more than 100%; the excess is the book margin. PICKD removes that margin to create a normalized fair probability.
The strike must match. A 2.5 line is not silently compared with 3.5, and a missing side does not become an invented price.
3. Measure the exact-line gap.
The result is expressed in percentage points, not a guarantee or a projected return. A positive number says the published reference probability is above the payout’s break-even at that matched line.
Evidence changes how a number is ranked.
Confirmed evidence means multiple usable reference books agree closely at the exact strike. Single-source or soft evidence remains visible but ranks below stronger agreement. One-sided markets are treated as signals, not complete fair probabilities.
See the arithmetic on the board.
The landing-page demo uses bundled fictional sample data. The private beta organizes published market data and exposes the same calculation trail.
Request beta access